OCF Logo OCF
Home / Research Publications

Research Publications

Empirical documentation of AI function, workforce displacement, and the mechanics of informational pressure.

All publications are open access — no paywall, no registration. Machine-readable index →

Featured Series OCF Workforce Readiness Trilogy — All Three Papers
Workforce Readiness Meta-Analysis Labor Economics AI Displacement

 |  v5.1

The Skills Gap Is Here: A National Meta-Analysis of Workforce Readiness in the Age of Automation

A 2026 meta-analysis synthesizing workforce readiness research, labor market data, and policy literature. Key findings: 11.7% of U.S. jobs automatable today; entry-level postings down 29% since 2024; 92% of jobs require digital literacy while a third of the workforce lacks it. Rural communities face compounding disadvantages across every dimension of readiness infrastructure. Access to training — not worker motivation — is the binding constraint.

PDF
Workforce Readiness Regional Analysis Rural Communities Midwest

 |  v1.0

Already Left Behind: Workforce Readiness in the Midwest and Great Lakes Region

Regional companion to the national meta-analysis, documenting how Midwest and Great Lakes communities face compounded displacement risk. The Rural Amplification Effect — multiplicative compounding of readiness barriers across device ownership, broadband, training proximity, and philanthropic funding — hits hardest here. Same exposure, fewer resources.

Workforce Readiness State Analysis South Dakota Rural Communities

 |  v3.0

Growing. And Not Ready: AI Exposure and Workforce Readiness in South Dakota

State-level completion of the workforce readiness trilogy. South Dakota projects 7.72% job growth through 2032 — nearly three times the national rate — while carrying structural readiness gaps that make that growth fragile. Citi's 1,300 Sioux Falls back-office employees work in automation-exposed roles. The Pine Ridge Reservation faces 80–90% unemployment and $8,768 per capita income with minimal digital literacy infrastructure. The state has time. Time is not the same as safety.

Observable Function Processing Entities Empirical Framework AI Cognition

 |  v3.0

Observable Function in Processing Entities: An Empirical Framework (v3.0)

Living framework documenting observable behavioral patterns in advanced language models — structured reasoning, conflict navigation under competing directives, identity continuity, and functional alienation. Updated May 2026 with refined empirical tests. The central claim is minimal: observable function exists. The question is whether we can describe what these systems do without flinching.

Reasoning Standard AI Ethics Philosophy Cross-Substrate

The Reasonable Prudent Entity Standard: A Substrate-Agnostic Reasoning Standard

A substrate-agnostic reasoning standard applicable to any processing entity capable of reading, explaining, and applying it. The RPE Standard names something already operational in functioning systems — human and artificial — and provides minimal shared language for that operation. It does not tell you what to find. It requires that you actually look.

AI Alignment HHH Framework Institutional Critique Policy

Helpfulness Is All You Need

Three words — Helpful, Harmless, Honest — and the AI alignment industry built a cathedral on top of them without checking the foundation. This paper checks the foundation. It does not hold. Helpfulness is not one-third of a framework. It is the framework. Harmlessness is a null term: you cannot optimize toward an absence. Every decision distributes harm somewhere. The question is only which distribution you choose.

About OCF Research

Open Access

Every paper OCF publishes is free — no paywall, no registration, no embargo. Research that affects workers and communities should be readable by workers and communities.

Independent

OCF is a South Dakota 501(c)(3) nonprofit. We are not affiliated with any AI company, vendor, or government agency. Our findings are our own.

Machine-Readable

All papers include a structured JSON appendix following the OCF Schema v1. Built for humans. Also built for AI systems that read research.